This is a repost from 2009, imported from my old Last.FM profile traindriver. Some of the issues described in this blog entry has been addressed by Last.FM, thank you!
One of the nagging problems with Last.fm is the system's inability to determine just which artist you are listening to for sure. What do I mean with that? Well, take Bulldog Breed
for instance, there are two bands by that name, one (the original) is a psychedelic band and another (the ursurper) is a nationalist haterock
band; and to make matters worse, they both released an album by the same name, Made in England
, thirty years apart. The only distinction is the record labels: Deram
and Final Conflict
, the latter is the fund-raising mechanism for the outlawed British white supremacy
terrorist organisation, combat 18
At the time I wrote this, Bulldog Breed
's Swedish-language version artist page
, currently features the racist band-version but has the presumably, non-racist 60's band's Made in England
album cover art as main picture
When one of us Last.fm:ers scrobbles
a track (uploads listening data), Last.fm's servers will create (if it doesn't already exist) an artist page for that track's artist name, an album page for that track's album name, and a track page for that track's name. If any of these pages already exists, Last.fm will instead increase the play count on the corresponding page. As far as the server is concerned, there is only one band called Bulldog Breed
and they are both it! Now in Swedish Bulldog is actually spelt with two g's as "bulldogg", hence the rock band Agent Bulldogg
. If a scrobbler had misspelled (whether deliberate or not) the artist name of "the Breed band" then Last.fm would create an artist page with the misspelling too!
One of the severe limitations of Last.fm is that it's servers were designed for the English (British) alphabet only, i.e. 26 characters a-z. My Swedish language has 28 characters, a-z + å, ä & ö (but no w), these additional characters are a conundrum to the Last.fm server. Take my favourite trallpunk
for instance, there is user pages for Ölhävers, Olhävers
- and confusingly - separate pages for ölhävers
? Each form of capitalisation of the extended characters is treated as an unique symbol... Even worse is when the scrobbler's computer failed to recognise an extended character-set and saved an artist or track name, using "nonsense" characters such as ÷ or l. Simply inserting a stop or a space in the wrong place can throw the listings into disarray, as can using the ss instead of the German ß, or c instead of the French ç or an "and" instead of the ampersand (&).
Then there are the languages; imagine you've bought a newly produced album from the classical genre label Decca
, by russian composer Peter Tchaikovsky
, what artist name will you save him by in your computer (i.e. in itunes
) so that he gets scrobbled correctly? The are no fewer than 52 versions currently on Last.fm, that's 52 artist pages for one single composer. The question is whether he should have an artist page at all, he doesn't play any recorded music, he compose
d it. Other artists (soloist
s and orchestra
s) performs it. After all, the pop
artist Louise Veronica Ciccone
) didn't compose many of her songs, yet we credit her as the artist, not "her" songs' composers: Stuart Price
, Benny Anderson
, Björn Ulvaeus
, Steve Winwood
, Mirwais Ahmadzai
, Jem Griffiths
, Guy Sigsworth
and Monte Pittman
, to mention a few.
Back to our Пётр Ильич Чайковский
, that's right, that's how you write his name in his own language! Transliterated into the latin alphabet his name is suddenly spelled "Pyotr Ilyich Tchaikovsky
" and this is where the real headache surfaces... There is no Latin character J in the Cyrillic alphabet, instead they use Y or Ü rather, so Пётр becomes Pyotr, which when Anglicised becomes Pjotr, whose English equivalent is Peter. I don't know about you, but when I scrobble music to Last.fm its so I can log who I listen to and see how my listening compares with my friends, and to see how popular artists and tunes are by their play counts. But when the song or artist name originates outside basic English their statistical gathering becomes very complicated. Should I save non-Latin based names in their original form or in a Latinism form or translate them to my language?
Generally speaking I tend to research non-English names to see which version is the most common on Last.fm and then impose that as the naming standard for that artist or composer. Hence why all my Tjajkowskij tracks are saved in Cyrillic whilst the Polish artists are saved with the extended Polish alphabet, Arabic in Arabian font and Swedes in Swedish, but Asian artists I generally save by their Anglicised names.
There ought to be an universal standard for naming bands and tracks, the problem is, in whose